Samsung to exit in-house automotive AP, focus on high-margin chips

Samsung's exit from automotive AP is not a cost-cutting reflex—it is a structural admission that full-stack coverage in System LSI has failed to generate returns. The technical ripple runs deeper than the design P&L. Automotive AP is a design-fab closed loop; pulling out strips Samsung Foundry of internal anchor volume. More critically, betting on CIS means staking on the perception layer of ADAS, which is being compressed by algorithmic efficiency. The decision layer (compute) is where durable margin lives. By 2027, as L3+ penetration accelerates, this trade-off inverts. On competition, Qualcomm's Snapdragon Ride and NXP's S32 will absorb the Tier 1 pipeline. The real wildcard, however, is MediaTek—leveraging mobile SoC DNA to undercut in automotive cockpit. Samsung doubling down on mobile SoC is playing in MediaTek's home turf, not opening a new front. Compliance and supply-chain: AEC-Q100 qualification cycles run 18-24 months. Samsung's departure effectively closes the 2026-plus new-platform window. For European OEMs, losing a non-Taiwan-China, non-US supplier reduces structural redundancy—a shift Brussels will scrutinize under the EU Chips Act. 12-24 month tail: Samsung Foundry's automotive design wins migrate toward TSMC and GlobalFoundries; CIS pricing resets on automotive volume; and Exynos' high-margin thesis collapses if it cannot establish a 2nm generational gap over TSMC-sourced rivals.